> Alexandria: the library your agents read

A curated, machine-readable index of papers, repos, and posts on building, evaluating,
and running AI agents. Index-only: we link to canonical sources and describe them;
we never copy or host third-party content. v0 public preview.

Facts: 74 seed entries · 24 topics · 9 use-case nodes · 3 kinds (paper, repo, post).
Records follow schema v1 (id, kind, title, url, authors, date, tags, license_note,
description, use_cases, source{type, fetched_at, via}).

Tools (contract sketch, implemented by the v0.2 MCP server):
- search(query, topic?, kind?, limit?) — vector + tag search over the corpus
- get_resource(id) — one full schema v1 record by stable id
- list_topics() — the taxonomy with live counts

Pages:
- [Home](index.html): what Alexandria is, how to point an agent here
- [Browse](browse.html): the taxonomy and featured entries
- [Use-cases](use-case.html): 9 job-first paths through the graph
- [Entry](entry.html): one resource record, human view and machine view
- [How it works](how-it-works.html): the curation pipeline and guardrails
- [API docs](api.html): the three-tool contract sketch

Policies: agents see no ads · sponsorship never outranks the graph · no standing
credentials · free unauthenticated reads when public.

Contact: VDR — df@vdrtech.net.